knaaptime
commented
4 years ago we removed the store_rasters function to reduce the dependency footprint, but we should still document how users [who have quilt] can quickly download nlcd data from our bucket with
import quilt3
p = quilt3.Package.browse("rasters/nlcd", "s3://spatial-ucr")
p["nlcd_2011.tif"].fetch()after some testing, it turns out theres a limitation in rasterstats that prevents us from leveraging quilt's streaming, and since basically all store_rasters did was wrap the above code, it makes more sense all around to just demo that code
Currently, some notebooks suggest downloading raster files similarly explained in https://github.com/locationtech-labs/geopyspark. However, since we have now the
data.store_rasters()function, we can leverage the usage of the package showing how to use this function rather than the current approach.